Honored to have this published in the Journal of Glaciology:

Rick Thoman@alaskawx.bsky.social · 2w ago

The late, great Carl Benson's final paper, "Revisiting the July 1954 melt event in Northwest Greenland" is now published. It's apt that his last paper is about his first science love: #Greenland #glaciology. @ruthmottram.eurosky.social www.cambridge.org/core/journal...

Exciting cutting-edge talks this afternoon with observations from under Thwaites Glacier taken with Icefin robot (Britney Schmidt & Peter Washam) - what's going on under there?! You'll want to see the recordings!! (Posted soon on YouTube & shared with IGS members - www.igsoc.org/membership) #igsoc

Jamin Greenbaum (@scrippsocean.bsky.social) Is subglacial freshwater in the fractured Thwaites Glacier Tongue? Ocean profilers / samples from 🚁. Great int'l coop - USA researchers/tech on Korean icebreaker! The answer: yes! Causing ice shelf breakdown. Very diff than outer ocean. #igsoc

Kirsty Tinto (@lamont.columbia.edu) on "Bathymetry of the Ross Ice Shelf at the TransAntarctic Mountains Front" - tectonics + erosion -> improving understanding of sub-ice cavities, to better know what/how melt may happen at critical locations for stability. #igsoc

Emma Cameron on "Depth-dependent influence of Greenland’s fjords on predicted submarine melt rates" - when extrapolating offshore thermal forcing into fjords, above 200m use a fjord model or near-glacier obs, but below 200m offshore water properties are more robust (in the absence of sills)! #igsoc

Shin Sugiyama on "Subglacial water pressure and its control on the dynamics of Glaciar Pio XI, Southern Patagonia Icefield" multiple boreholes to the bed, speed variations correlated with water pressure - but everywhere, showing variable subglacial conditions related to melt, rain & more #igsoc

Duncan Wheeler on "Observations of a Weakly Buoyant Plume at a Tidewater Glacier Suggest Englacial Discharge" - These observations reveal a new type of buoyant plume not accounted for by traditional theories #igsoc

Jesse Cusack on "Observed ocean current variability at a marine-terminating glacier" - Locally generated waves can help address melt discrepancies, need to study ice-melt by unsteady processes, calving effects need more study, and there is a continuum of systems with different local forcings #igsoc

Ginny Catania sharing some really cool observations of Greenland glacier terminus using Nereid Under Ice submersible - sediments, structures, melt, and more. (Related: her IGS seminar talk from earlier this year www.youtube.com/watch?v=EmK7...) #igsoc

29 Apr 2026 - Ginny Catania: Understanding Terminus Behavior Using Machine Learning & In-Situ Obs...

YouTube video by International Glaciological Society

youtube.com

Our Ice-Ocean symposium is kicking off in La Jolla, California! Bringing together over 100 scientists to share results and discuss this urgent are for the future of glaciers and ice sheets. We'll try to share some highlights here over the week. igs2026.ucsd.edu #igsoc

Thanks to Dartmouth for this interview with @mathieu-ice.bsky.social - lead for the AI in Glaciology Symposium, about the meeting, what it meant for the AI & glaciology research communities, where research is going & why it’s an exciting time to be a student!! home.dartmouth.edu/news/2026/07...

AI's Growing Role in Glaciology | Dartmouth

home.dartmouth.edu

IGS was proud to co-sponsor the Karthaus 2026 Summer School on Ice Sheets and Glaciers in the Climate System. You can see this year's meeting report - and if you're a student, maybe consider applying in the future! www.igsoc.org/karthaus2026

Meeting Report: Karthaus 2026 Summer School on Ice Sheets and Glaciers in the Climate System | IGS

Karthaus 2026 participants learning about the Lazaun rock glacier and its environment. On 18 May 2026, 36 students (mainly Ph.D.) and 10 teachers arrived

igsoc.org

Marlena Reil sharing SnowGalileo: A multi-sensor transformer for snow cover mapping in mountainous regions. Successful data fusion method for daily, high-resolution snow cover mapping. Still working on limitations like forest canopy and better validation regionally & temporally. #igsoc

.@colinrmeyer.bsky.social: FirnLearn, a neural network firn model. Trained separately for different regions. In Greenland, especially, it does quite well & better than current model (which works oddly well). Improves with more observations - and could be used to drive future sampling, too? #igsoc

Karina Zikan: autoGRITTI - generating Greenland Ice Sheet Land Terminus positions. Object-oriented random forest classifications in GEE, interpreted using physical info. 2022 completed, 2020 in progress. Handles moraines/debris, ice melange & lakes. ML is a starting point. Automated in reach! #igsoc

Bryony Freer (@schmidtsciences.bsky.social AI Fellow) looking at Antarctic tidal grounding zone dynamics. Using ICESat-2 to look at inland limit of tidal flexure. Critical boundary condition, trying a bayesian inversion model to scale up to the whole continent, but still work to do. #igsoc

Omar Faruque: Causal inference for analyzing drivers of GrIS melt. Air temp & precip > various melt impacts. Large-scale variability (NAO) reduces clouds and increase melt. From 9 causal links, 4 agree with domain knowledge at some time lag, but 5 don't. Need to consider more feedback loops. #igsoc

Jonathan Ryan is using a deep learning model with @planet.com data to classify melt features on the Greenland Ice Sheet. 97% accuracy with a lightweight U-Net mode, robust across different imagery processing levels & better than global thresholding esp with lower imagery quality). #igsoc

Maya Maciel-Seidman models Greenland runoff using Transfer Learning. Trained on MAR at emulator stage & then fine-tuned using in-situ runoff. Albedo is an important input. TL is most effective in areas where MAR is most biased, but it may be overfit in certain catchments. More tuning data! #igsoc

Mashrekur Rahman & Aleah Sommers tag-teamed a presentation on subglacial hydrology & glacier sliding dynamics using machine learning, including for evaluation. Primarily interested in glacier velocity. Eventual goal is to move from expert-only simulations to community-useful knowledge! #igsoc

Max Filter presented a generative model for subglacial topography (from GlaMacLes 2025 - glamacles.github.io - you should apply for next year's summer school!!). Wants a realistic & scientifically-useful output. Can condition on known deglaciated / nearby terrain / radar & match flow. #igsoc

Niya Shao @gatorglaciology.bsky.social Combining geostatistics and physics-based mass conservation methods to improve bed topography representations using MCMC. #igsoc

Mansa Krishna on Taking Physics-Informed Neural-Networks to the Higher-Order, looking at bed topography under Greenland. Does a good job, AND also to include vertical shear rather than SSA to compare closer to BedMachine and get full-depth troughs. #igsoc

Machine-learned global glacier ice volume, Niccolò Maffezzoli - lower error than previous, inputs include thickness & 26 different predictors (not all independent). No physics, scary outliers & challenging areas of low quality DEM & inputs (e.g., Asia). But fast & the model learns globally!! #igsoc

We are back with a few talks on snow & sea ice this morning: Molly Balkan is connecting sea ice retreat and ice shelf melt, Sai Vikas Amaraneni is Predicting Antarctic Sea Ice with a deep learning model, and Ayush Prasad is emulating snow depth & density on sea ice with reasonable success! #igsoc

Save the Date - IGS Symposium on Snow in a Warming Climate: Processes, Extremes, and Disaster Risk (Aomori, 7-12 November 2027) | Share knowledge, foster collaboration & explore the latest developments in snow science! Full info, 1st circular & pre-registration form: www.igsoc.org/event/snow2027

Doug Brinkerhoff on emulators vs classical methods - a path to creating Glide: GPU-accelerated Ice Dynamics Engine. Reach out to Doug if this might be a useful tool for you! #igsoc

After lunch, we are back! Cheng Gong is sharing PINNICLE (Physics-Informed Neural Networks for Ice and CLimatE) - a flexible framework to test and integrate physics with data. (IMHO: Some cool bedmap comparisons!) #igsoc

Our research group and @precise-project.bsky.social is represented by Clement Cherblanc who will be presenting the first results from our SMB emulator in Antarctica at the poster session. Looks like a very interesting line-up

International Glaciological Society@igsoc.bsky.social · last mo.

Our AI in Glaciology Symposium meeting has started! Hosted by @mathieu-ice.bsky.social & team at Dartmouth. We have been welcomed by local organizers, VPR Dean Madden & heard from NASA Program manager for Cryospheric Science Thorsten Markus -- AI is the future of the cryosphere. #igsoc

Tancrède Leger on using IGM to model past icefields, mainly in the European alps - looking at ice extent & thickness, as well as using particle tracking to understand debris transport with 3D particle transport, too! #igsoc